Managing Dynamic Events in Full-Load States of Semiconductor Manufacturing Chains

Semiconductor manufacturing is a complicated operation that is plagued by all sorts of uncertainties in process yield, equipment reliability, IC demands and product mixes. This paper describes a new proactive planning method for responding to dynamic events that pose a threat to degenerate the supply services of manufacturing chains. A dynamic system model for manufacturing chains under exogenous demand shocks is first described. As the impact of dynamic events is the most severe when manufacturing chains are in the full-load state, a procedure is next presented for constructing full-load production functions of flexible capacity. The third part of this paper develops an optimal control model for integrating the production functions of multiple production units. Given an inordinary event of demand shock, this model can be used to determine if the shock can be absorbed by the manufacturing chains without degrading its fulfilment services. Numerical examples are given to demonstrate the working of the method.